Characterization of fractured reservoir using offset VSPs : case study from the Varandei field, Northern Russia

Anatoli A. Tikhonov (PetroAlliance), Ludmila A. Plekhodkina (Lukoil), and Enru Liu (British Geological Survey) in this technical article describe a method for the delineation of fracture-induced azimuthal anisotropic intervals in the subsurface supported by a case study with Russian data. Traditional seismic techniques of examining fracture-induced seismic anisotropy in a borehole utilize either walkaround observations of direct compressional waves (Horne 2003; Horne et al., 2002; Thompson et al., 2002), observations of vertical shear-waves with different polarizations (Brodov et al., 1991, Liu, Crampin and Queen, 1991; Winterstein, Ge, and Meadows 2001), or converted PS waves in offset VSP data (Horne et al., 2002). The walkaround VSP data allows the construction of the P-wave velocities (Vp) indicatrix. The observations of vertical shear-waves are based on shear-wave splitting or birefringence measurements. The elastic moduli are usually calculated from the kinematic parameters of seismic waves assuming the model of transversally anisotropic media with a horizontal symmetry axis. For more complicated models (for example, thin-layer section with vertical fractures leading to orthorhombic model) the 3C multi-azimuth, multi-offset VSP is used (Dewangan and Grechka, 2002). Our goal is to extract information about seismic anisotropy due to the vertical fracturing from routinely recorded offset VSP data obtained with non-directional or explosive sources. We attempt to answer the following two questions: (1) Is it possible to control the polarization of converted PS waves in order to obtain pure S1 and S2 modes? and (2), How to remove the influence of offset PS ray paths in order to estimate time delay between S1 and S2 waves? The idea of controlling the polarizations of converted down-going PS waves is based on the use of linear combinations of wavefields recorded along a vertical seismic profile from pairs of nondirectional sources located in different directions (Figure 1). We assume that the shear wave polarizations after P to S mode conversion at a given interface in horizontally layered isotropic media is contained in a vertical ray plane passing through the sources and receivers. The original technique, which uses the rotation of shear-wave polarizations proposed by Brodov et al. (1991), assumes that two different shear modes propagate along the same ray paths. However, surface seismic data use the rotation method to detect the directions of anisotropy symmetry axes using data obtained along different ray paths. This technique assumes that NMO corrections reduce different ray trajectories to the same vertical ray. We use a ray-tracing technique to remove the influence of PS ray geometry on arrival times and to reduce to vertical profile the wavefield registered from offset VSP. The initial kinematic model is built by the inversion of direct P and converted PS wave travel-times. We correct the 2D kinematic model after imaging of depth migrated seismic section